Implementation:Openai Openai node Jobs Create

Knowledge Sources	openai-node OpenAI Fine-tuning API
Domains	Fine_Tuning, Model_Training
Last Updated	2026-02-15 00:00 GMT

Overview

Concrete tool for creating fine-tuning jobs on the OpenAI platform provided by the openai-node SDK.

Description

The Jobs.create() method sends a POST to /fine_tuning/jobs to initiate model training. It accepts a JobCreateParams object specifying the training file, base model, optional validation file, training method (supervised, DPO, or reinforcement), and hyperparameters.

Usage

Call this after uploading and verifying training data. The returned job object contains an ID for monitoring progress.

Code Reference

Source Location

• Repository: openai-node
• File: src/resources/fine-tuning/jobs/jobs.ts
• Lines: L38-40 (create method), L429-620 (JobCreateParams)

Signature

class Jobs extends APIResource {
  create(
    body: JobCreateParams,
    options?: RequestOptions,
  ): APIPromise<FineTuningJob>;
}

interface JobCreateParams {
  model: string;               // Base model (e.g., 'gpt-4o-mini')
  training_file: string;       // File ID from upload
  validation_file?: string;    // Optional validation file ID
  suffix?: string;             // Model name suffix
  seed?: number;               // Reproducibility seed
  method?: {
    type: 'supervised' | 'dpo' | 'reinforcement';
    supervised?: SupervisedMethod;
    dpo?: DpoMethod;
    reinforcement?: ReinforcementMethod;
  };
}

Import

import OpenAI from 'openai';
// Access via: client.fineTuning.jobs.create(...)

I/O Contract

Inputs

Name	Type	Required	Description
model	string	Yes	Base model to fine-tune
training_file	string	Yes	Training file ID
validation_file	string	No	Validation file ID for eval metrics
suffix	string	No	Custom suffix for fine-tuned model name
method	object	No	Training method (supervised, dpo, reinforcement) with hyperparameters
seed	number	No	Random seed for reproducibility

Outputs

Name	Type	Description
job	FineTuningJob	Job object with id, model, status, training_file, created_at, fine_tuned_model (null until complete)

Usage Examples

import OpenAI from 'openai';

const client = new OpenAI();

const job = await client.fineTuning.jobs.create({
  model: 'gpt-4o-mini-2024-07-18',
  training_file: 'file-abc123',
  suffix: 'my-custom-model',
  method: {
    type: 'supervised',
    supervised: {
      hyperparameters: {
        n_epochs: 3,
      },
    },
  },
});

console.log('Job ID:', job.id);
console.log('Status:', job.status);

Related Pages

Implements Principle

• Principle:Openai_Openai_node_Fine_Tuning_Job_Creation

Requires Environment

• Environment:Openai_Openai_node_OpenAI_API_Credentials